alarsyo/spot
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

Reimplement basic_reduce()'s rules in ltl_simplifier.

Browse source 
So far I have only checked these rewritings with reduccmp.test.
There are probably a few kinks to iron out.

* src/ltlvisit/simplify.cc: Reimplement most of the basic
rewriting rules, leaving some FIXME comments for dubious ones.
* src/ltlast/multop.cc, src/ltlast/multop.hh: Ignore NULL
pointers in the vector.
* src/ltlvisit/reduce.cc (reduce): Do not call basic_reduce().
* src/ltltest/reduccmp.test: Adjust tests.
This commit is contained in:
Alexandre Duret-Lutz 2011-08-22 19:34:09 +02:00
parent e3e0f913b6
commit ca2fe4f3f8
5 changed files with 1020 additions and 45 deletions
Download patch file Download diff file Expand all files Collapse all files

11
src/ltltest/reduccmp.test
View file

@ -100,8 +100,10 @@ for x in ../reduccmp ../reductaustr; do
run 0 $x '(a U b) | (a U c)' 'a U (b | c)'
run 0 $x '(a R b) | (c R b)' '(a | c) R b'
run 0 $x 'X(a & GFb)' 'Xa & GFb'
run 0 $x 'X(a | GFb)' 'Xa | GFb'
run 0 $x 'Xa & FGb' 'X(a & FGb)'
run 0 $x 'Xa | FGb' 'Xa | FGb' # We'd prefer 'X(a | FGb)'
run 0 $x 'Xa & GFb' 'Xa & GFb' # 'X(a & GFb)'
run 0 $x 'Xa | GFb' 'X(a | GFb)'
# The following is not reduced to F(a) & GFb. because
# (1) is does not help the translate the formula into a
# smaller automaton, and ...
@ -109,10 +111,9 @@ for x in ../reduccmp ../reductaustr; do
# (2) ... it would hinder this useful reduction (that helps to
# produce a smaller automaton)
run 0 $x 'F(f1 & GF(f2)) | F(a & GF(b))' 'F((f1&GFf2)|(a&GFb))'
# FIXME: Don't we want the opposite rewriting
run 0 $x 'Fa & GFb' 'Fa & GFb' # We'd prefer 'F(a & GFb)'
run 0 $x 'G(a | GFb)' 'Ga | GFb'
run 0 $x 'X(a & GFb & c)' 'X(a & c) & GFb'
run 0 $x 'X(a | GFb | c)' 'X(a | c) | GFb'
# The following is not reduced to F(a & c) & GF(b) for the same
# reason as above.
run 0 $x 'F(a & GFb & c)' 'F(a & GFb & c)'